Bones include many proteins, some of which induce or promote bone growth. A great deal of research has been directed to producing, either by recombinant DNA techniques or by purification of naturally occurring proteins, specific osteoinductive proteins. Such proteins and a variety of processes for obtaining them are the subject of numerous patents. However, very little work has been directed to the economic, large scale commercial production of useful osteoinductive proteins.
Collagen Corporation of Palo Alto, Calif. is the assignee of a number of patents directed to osteoinductive proteins U.S. Pat. No. 4,434,094 by Seyedin et al., issued Feb. 28, 1984 identifies a process to partially purify an osteogenic factor and isolate a non-fibrous protein having a molecular weight less than 30 kilodaltons (kD) from demineralized bone extract using cation exchange chromatography. A partially purified bone-inducing factor of 10 to 30 kD and the purification process including extraction from demineralized bone, gel filtration, and cation exchange chromatography on a carboxymethyl cellulose column, and which may include reverse phase-high performance liquid chromatography (HPLC), is described in U.S. Pat. No. 4,627,982 by Seyedin et al., issued Dec. 9, 1986. Also by Seyedin et al. and assigned to Collagen Corp., U.S. Pat. No. 4,774,228 issued Sep. 27, 1988, describes two 26 kD proteins found in bone having activity in a TGF-.beta. assay and purified using a process similar to that taught in Seyedin's U.S. Pat. No. '094 patent but including reverse phase HPLC or acetic acid-urea gel electrophoresis, where the purified proteins exhibit chondrogenic activity (purportedly related to bone formation). U.S. Pat. No. 4,863,732 by Nathan et al., issued Sep. 5, 1989 is directed to an injectable solution of an osteogenic factor such as that described in Seyedin's U.S. Pat. No. '982 patent, combined with atelopeptide collagen and further purified by coprecipitation. Other patents relate to mixtures of atelopeptide collagen material, e.g. U.S. Patent Nos. 4,789,663 by Wallace et al., issued Dec. 6, 1988 and U.S. Pat. No. 4,795,467 by Piez et al. issued Jan. 3, 1989.
Marshall R. Urist is an inventor named in numerous patents in the field of bone inducing agents. U.S. Pat. No. 4,294,753 by Urist, issued Oct. 13, 1981, describes a process for obtaining bone morphogenetic protein (BMP) by treating demineralized bone with a neutral salt to transform the bone collagen to gelatin, extracting the BMP with a solubilizing agent, then removing the solubilizing agent and salt by dialysis to precipitate the BMP. It is recognized that precipitation of proteins from solution is not highly selective. The BMP, with a molecular weight ranging from 1,000 to 100,000 is the subject of U.S. Pat. No. 4,455,256 by Urist, issued Jun. 19, 1984. U.S. Pat. No.4,619,989 by Urist, issued Oct. 28, 1986, discloses an improved process for further purifying and isolating human and bovine BMP compositions and factors, including additional dialysis and co-precipitation steps. U.S. Pat. No. 4,761,471 by Urist issued Aug. 2, 1988 relates to products obtained by the aforementioned process including a substantially pure BMP composition containing an active 17.5 kD (human) or 18.5 kD (bovine) BMP factor and BMP associated proteins with molecular weights of approximately 14, 22, 24, and 34 kD which may enhance but do not induce bone formation.
U.S. Pat. No. 4,877,864 by Wang, et al., issued Oct. 31, 1989 discloses human and bovine bone inductive factors of approximately 28 kD to 30 kD and characterized by a specific peptide sequence, which may be produced by recombinant gene techniques.
U.S. Pat. No. 4,804,744 by Sen, issued Feb. 14, 1989 identifies a primary osteogenic protein (P.sub.3) with a molecular weight of about 22 to 24 kD. This patent also identifies proteins P.sub.2 and P.sub.4 which are nonosteogenic without P.sub.3, and further identifies a method for isolating P.sub.3 from demineralized bone tissue including extractions, dialysis, gel filtration and HPLC steps.
As is apparent, it would be desirable to have a mixture of proteins which are highly osteoinductively active. It would be beneficial if such proteins could be produced efficiently and effectively on a commercial scale. It would also be beneficial if such proteins could be produced in a manner designed to minimize degradation of such proteins while maximizing production. It would also be beneficial if such proteins could be produced using relatively well-known unit operations in a process which is tolerant of minor variations in process conditions. It would also be beneficial if the mixture of proteins could be produced directly, without having to first obtain single specific proteins and remix them to attain the desired mixture.